Verified Learning for Compiler Optimization: An LLM-Guided Architecture with Formal Control

2026-09-22

AI Summary

研究使用大型语言模型结合形式化验证方法参与编译器优化，通过生成式重写与形式等价性检查耦合，确保转换的语义正确性和优化效果。

Abstract

Compiler optimizations traditionally rely on handcrafted heuristics that often fail to generalize across programs and architectures. We investigate whether large language models can participate in compiler optimization through a verification-centered systems architecture that couples generative rewriting with formal equivalence checking. Using lazification in LLVM IR as a case study, we fine-tune a code-centric LLM on transformations produced by Wyvern and embed Alive2 into a feedback loop that enforces semantic preservation for every generated rewrite. Correctness is enforced externally as a runtime control layer rather than learned implicitly. During inference, candidate transformations are symbolically validated and regenerated when necessary, ensuring accepted rewrites satisfy formal constraints. On the LLVM test suite, the fine-tuned model reproduces core optimization behaviors while applying fewer transformations overall. Although Wyvern remains faster on most benchmarks, 9.8% achieve comparable or improved runtime under the learned system, with no semantic violations observed. Verification overhead remains bounded and convergence stable. These results demonstrate that generative AI components can be safely integrated into compiler pipelines through deterministic validation and structured feedback, offering a scalable architectural pattern for trustworthy AI-driven software infrastructure.

Problem

• compiler optimization
• large language models
• formal verification
• generalization
• semantic preservation

Innovation

• large language models
• formal equivalence checking
• generative rewriting
• semantic preservation
• deterministic validation